Reading through the history of that is ridiculous.
E: Hey we're pretty sure this is going to kill people.
M: Are you sure? Maybe we can just do it while thinking really positive thoughts.
Meanwhile one of the managers responsible for giving the 'go signal' still blames the engineers. "Hey, if it was unsafe the engineers should have said so." Ignoring that he had been bullying them for hours to shut up about how it was unsafe.
The dramatization of the event in one of the documentaries stands out to me. "I'm going to go around the room and if anyone knows that this thing is definitely unsafe, then they need to speak up." WRONG. You go around the room and make everyone explicitly say it's safe. If anyone refuses you don't launch. They might seem like logical duals, but there is a huge difference.
Boeing thought their last launch delay was just going to take a small amount of time too. Then, as they dug into it, they basically put it on indefinite hold when they realized they had no idea on what to do about the issue causing the dealy.
Of course they will come out and say "just a short delay" immediately, but who knows what the actual delay gets to when they really start looking at it. Besides, how can they really know that it is okay without taking it down and putting it through all of the motions? A bent arm in a manner not visible while folded in launch config but prevents it from unfolding after launch is a very dumb reason to not have this thing work.
Maybe, just maybe, if we stopped building telescopes like fragile porcelain dolls with glass heads that can't take a scratch and started building them like tanks that can take a beating..
Soon we might have an option to avoid the extreme mass and size restrictions caused by having to squeeze the JWT on top of Ariane 5 - if/when Starship proves viable, then we could probably cut the cost of the telescope development by an order of magnitude by being able to choose a reasonable tradeoff instead of sacrificing everything (complexity, cost, reliability) just to cut down some weight and size.
Its about balancing between the needed properties, strength and weight. Similar thing for airplanes, a minor collision can ground a plane but would be a non issue for a tank.
>”Maybe, just maybe, if we stopped building telescopes like fragile porcelain dolls…”
It’s a pet peeve of mine but whenever I read “Maybe, just maybe” I instantly dislike whatever comes next. It’s an indicator that the post is sardonic and dramatically over simplified.
The real reason why the JWST is so delicate is because it’s a large satellite that had to be crammed into a tiny fairing. That means it has to be folded up like the worlds most intricate origami. All those moving parts, coupled with incredibly thin and weight saving components means that deploying it successfully is tremendously fraught with complexity. If we had a wider diameter rocket, JWST wouldn’t be nearly as complicated. But that simply wasn’t the case in 1997 when it was first designed.
"Technicians were preparing to attach Webb to the launch vehicle adapter, which is used to integrate the observatory with the upper stage of the Ariane 5 rocket," NASA said in a blog post. "A sudden, unplanned release of a clamp band—which secures Webb to the launch vehicle adapter—caused a vibration throughout the observatory."
Something about a rocket having a user's manual makes me smile. Seems like a proper space agency name for this document would be much more space program sounding than just "user's manual". Even software programmers use terms like User Acceptance Training instead of just training. That one always makes me laugh too though, "You're going to learn to accept this software whether you like it or not."
I have never heard of User Acceptance Training. I have heard of User Acceptance Testing (UAT), which is about users testing the software, not training users on the software. The goal of UAT is to work with the user to verify that the software's functionality and usability are "acceptable". https://en.wikipedia.org/wiki/Acceptance_testing#User_accept...
It's always agrressive against the users. The users are the poor saps that have to accept whatever the devs have droped on them. The users have no say other than suggest it doesn't work, doesn't promote efficient workflows, whatever. It's then up to the devs to fix, but they feel like they've delivered according to specs so tough. The User will Accept Whatever from the all powerful dev team. How about User Acceptability Testing. At least it then makes it sound like the devs are going to give a damn about the users feedback.
Is this a Word Autosave, or someone aactually hitting save but couldn't be bothered to actually name the doc? It's clearly Word generating the filename.
I think it just goes to show that nobody likes creating documentation.
But it is Word that takes the firsr line of text to use as the filename, not the PDF writer. It could be a combo though. Word using the first line of text for the filename, then the PDF export appending the name of the app.
> This user’s guide is intended for pre-contract mission planning and for understanding SpaceX’s standard services. The user’s guide is not intended for detailed design use. Data for detailed design purposes will be exchanged directly between a SpaceX customer and a SpaceX mission manager.
Engineers at heart wouldn't let a single point of failure in in the design of it to let a pin drop unexpectedly. Is the vehicle insured for a do-over should it fail to launch unspectacularly and land in the water?
After watching an official JWST video https://www.youtube.com/watch?v=uUAvXYW5bmI, they seem almost proud of having several hundred of single points of failure and an immensely complex deployment process that takes several weeks and must work perfectly. I don't know what engineer could be proud of coming up with a design like that.
Its a machine that is extremely mass limited, has many different groups of people influencing different systems, and is extremely complex.
If the HARD requirement is "you have this much mass and volume to make a thing to a particular confidence score" you make the damn thing and hope you are on the correct side of the dice.
No single engineer gives a shit about the entire system because no single person is equipped with the human ability of comprehending each individual part that has whatever confidence rating that is required.
Single points of failure are inevitable. engineers could always make a copy of the entire design but that would double the costs and you would still be left with unknown points of failure that the telescope double would have.
The solar shield is like rolling out five layers of tightly packed tinfoil from several points without tearing. Claiming that you would do away with single points of failure if you were the one designing it is too easy to say on HN
I'm not arguing you could completely go away with single points of failure when designing a spacecraft while being limited by the payload mass and volume. But the whole JWST unfolding sequence looks like a way too risky gamble to me. Maybe, it was worthwhile to sacrifice some of the scientific payload mass to add more redundancies into the solar shield design and increase the chances the telescope would be able to reach the beginning of its observation process.
There's also the whole "what if we could avoid mass and volume constraints" tangent, that was discussed in the yesterday's Starship article, but it's more of a theoretical consideration for now.
Honestly, I thought the same about the whole Perseverance landing sequence on Mars. But now it's done, and the rover is running around doing its thing.
So, what I learned is that single points of failure on initial deployment -- which by definition only has to run once -- is a drastically different cost vs risk calculation than something that happens during normal operations.
The unfolding sequence looked complicated to me, too, and then I remembered I am just a guy online.
You are right that a magical rocket could solve many of NASAs problems (just launch ten Webbs and use the ones that don't work as sunshades for the rest!) but I think NASA is being pretty practical in not waiting for one to be invented.
What was the original budget for JWST? Where's it at now? From your assessment and the numbers, it sounds like there's not just double but triple redundancy now.
In 1996 when it was a proposed idea the estimated cost was 500 million USD and the proposed launch was 2007. But the design kept getting more and more lofty and the project got more and more complex. The initial estimate was a bit off, it has cost about 9.6 billion USD so far.
Still though, if everything works and we learn more about the universe, I think it'll have been worth it.
The question is what engineer could be proud of finding a solution within such difficult constraints, and the answer is a really good bunch of engineers.
Very aggravating to hear them use units like "thousandths of an inch thick" and "feet of cables". Does NASA really still work with imperial units of measure?
The mass constraint in space engineering is a real killer, unfortunately. I doubt they want so many single points of failure, but I also understand it can be effectively impossible to add redundancy within mass window.
These people aren’t dumb. I am pretty sure they went through a lot of options before settling on a design. Keep in mind that a lot of space devices are outdated by the time they are being deployed.
I'd imagine there were redundant systems that caught it, but it still caused a significant shift in load distribution and possibly the satellite to drop some distance which would impart an acceleration load.
I wonder how much shock the system is going to take to use a persuasion tool (aka mallet) to release it from this "redundant" system you imagine. My imagination is much less prolific and see it wedged in at an angle in a way not imagined by the deigners.
In many systems, yes, that's to be avoided. However, the space industry as a whole is so tightly constrained that it's a necessary evil to eliminate redundancy and even reduce safety factors below ideal.
Almost no government space project is insured. The govt is self-insured, generally. Which incidentally, means private space companies are effectively out-sourcing much of their govt contract risk to the taxpayer
What is the plan for these sorts of highly-expensive, super-delicate instruments in an environment where Russia (or other spacefaring nations) just blow things up leaving huge clouds of debris that would destroy it in seconds?
Does the JWST just operate at a different level of orbit or something?
Yup - JWST will be way high up in space, much farther away than the Moon and plenty far from the debris we talk about in low Earth orbit. The telescope actually orbits the Sun, not Earth, at a point called "Earth-Sun L2", which is a point in line with the Earth. L2 is one of 5 Lagrange points in the Earth-Sun system, which are special spots objects can orbit around (stably or unstably) without there being a central body.
Sorry, should have made my comment more clear. Wasn't suggesting that Russia would blow up JWST. Rather, it might inadvertently be damaged by one destroying a different satellite, and the debris from that colliding with JWST.
Maybe you should name and shame everyone that done this same thing, even if you don't remember similar actions the comments on the Russia storied remind us that other assholes also did the same, some even blew up nukes in orbit causing malfunction to satellites.
Yeah, it is depression to think that lot of things are at the mercy of a few giants whims.
It's worth noting that Russia is not the only country that has done this. The US, China, and India have all blown up a satellite with a rocket from Earth.
Here a couple of Wall Street journal commentary (AKA opinion) pieces on the topic. They all talk about different aspects of it, the first being probably the most interesting around your point of view.
Worth noting, indeed; thank you. Probably should have worded my comment more carefully to just be generally worried about debris in Earth orbits no matter the source and not singling out Russia.
In case you didn't actually click any of the articles, I recommend at least reading the first. It talks about needing a new treaty for space to deal with space debris. The other two focus on national defense from a US perspective.
I think the difference is whether the vibration is expected.
Vibrations from being launched are probably accounted for and tested numerous times, while an unintended vibration caused by a "sudden, unplanned release of a clamp band" could cause unforeseen damages to the instrument.
The statement does not actually say anything about the magnitude of the vibrations during this incident, let alone how they compare to those expected during launch.
"Some vibration" doesn't give the same feeling in the reader as the techs and engineers on the floor felt when they heard a 10 Bn USD science project that millions of people are hoping to learn from crashes into the ground.
We should be launching a new JWST every year. And we could, for a tiny fraction of the cost of a single large DOD project. Hopefully Starship will make these once-in-a-generation level missions more commonplace, so we're not losing our minds over every tiny thing that can go wrong with this thing.
Are you implying that yearly launches would reduce cost through economies of scale or something? Or just that JWST costs less than the F-35 program?
While having a Starship sized fairing to put JWST in for launch would probably reduce some packaging complexity, I'm not sure the launch vehicle is the major source of cost in this case.
Does anyone know how much of the price tag for JWST is manufacturing, as opposed to research/design/development costs? I would expect the actual manufacturing is a very small fraction, so I don't understand why they haven't built two of them - would seem to be good insurance for just this sort of mishap. And if you do find a problem when the first one is commissioned at the inaccessible L2 point, you at least have some chance to correct it on the second one. Best case is you get two for nearly the price of one.
They may not have the labor to build two of them. It's exceptionally difficult to put together the team necessary to build one of them.
The US isn't a serious nation any longer. The US is increasingly becoming a clown show. The response you'd get if you suggested to consider building two of them - for the reasons you mention - is that people would freak out (two boondoggles, $20 billion in cost, decades of time; that would be the flailing, screeching response).
What's the ROI? The hard limit for stuff which is not aimed at producing an improvement in quality of life of taxpayers should be in the dozens of millions , not billions.
Quality of life improvement can be delayed 40 or 50 years, but there should be at least a path.
James Webb Space Telescope has no path, nor does Hubble, nor does the ISS, nor do the Mars missions (private or public), and neither did the Apollo missions as the program was shut down for bankruptcy essentially.
Had the Soviet done the same we'd be laughing at their juvinile understanding of economics, ROI and ROE.
On the other hand NASA mission to collide with an asteroid has a path to save us all , ITER has a path (no pun intended) , mRNA research has a freaking highway at this point, biotech also has a path, even China's bridges to nowhere to pump GDP have a path sometime in the future.
Sounds like a good argument to severely cut millitary spending, not science spending. Also, pure science is the part of eventually getting useful qol improvements that absolutely cannot be done through simple short sighted private research, but proves time and time again to be invaluable.
> Also, pure science is the part of eventually getting useful qol improvements that absolutely cannot be done through simple short sighted private research, but proves time and time again to be invaluable
Seems like a good pitch to fund National Labs instead.
National Labs != space.
Unless you have a clear and defined QOL plan then space has proven to be a money sink for the longest time.
The Apollo mission didn't have any clear plan, the Voyager-1-2 missions didn't have any clear plan, Mars missions won't have any clear plan, the ISS didn't have any clear plan, the space shuttle didn't have any clear plan.
Matter of fact it's the opposite of what you claim, if you leave space in the hands of military maybe some QOL trickle down will happen (GPS, Sat communications, ICBMs keeping us safe).
If you leave it in the hands of wide eyed individuals who want to discover exoplanets and clusters which we'll never reach then you'll just throw money down the sink.
Earth is going to get very crowded in the next couple of centuries. The economic case for large-scale expansion into space isn't exactly clear at this point, but when things get tight I'd rather have state and corporate actors looking outwards for their next growth fix, and for that to be a possibility we need to be investing in space technology now even if there is no tangible short-term ROI.
Unless we can change our essential nature, I don't see any sustainable path where human civilization stays confined to Earth forever. Those doomsday weapons "keeping us safe" will inevitably be used, especially when it becomes clear that we're utterly trapped.
You probably make a valid comment about USA, but I would hazard to bet that it is, in fact, the zeitgeist/ cultural / societal climate of the world in general has been stripped of rationality in favour of clickbait, sensationalism and dare I say, value faking (I support X because it makes me a appear to be a better person, but in reality I don't authentically live.). [There is a term for this that I can't think of right now.]
The root problem with JWST was not being able to get the money in the first place. That led to budgeting shenanigans responsible for much of the project delay. The single telescope came within a whisker of being scrapped by Congress.
If you gave projects like Webb the kind of budget that now gets eaten by human space flight, you could have telescopes filling half the sky. But NASA has to work within the really severe funding constraints imposed by the white elephant that is ISS.
I don't think there's a meaningful distinction between "manufacturing" and other development activities in this case. This sort of thing, an effort to build a single example of a device using a mountain of new technology, just can't follow the process of design the whole thing, then manufacture it, then we're done. Kind of like waterfall vs agile software development. They're probably constantly fabricating and testing subsystems in various ways, iterating on manufacturing technique to get quality good enough, making changes in other related subsystems as a result of things they found to be impractical with the first, making sure subsystems actually integrate with each other as expected, etc.
To everyone saying "what's the big deal, it sees more vibration on launch":
It sounds to me like this is more of an unintended shock event (see zeeb's response below) that caused an oscillation at the natural frequency ("vibration") but they tortuously avoided using the word "shock".
This would result in a given shock response spectrum [1], which effectively can amplify the input if it hits any part's natural frequency - e.g. a 1G input shock could become a 5G response somewhere in the satellite. I assume this wasn't accounted for in design so they'll determine if this could have damaged any onboard systems.
Launch vibration has a well known Power Spectral Density [2] which effectively says how much energy is being input at every frequency. If they design the system so no parts have a natural frequency at a high-energy peak it's safe.
Shock and vibration are closely related, but very different in practice. It's all about natural frequencies and what frequencies the energy is being input at.
Is the JWST in a state where they can open it up and inspect? Or are they just going to run simulations to see if the shock/vibration would cause any anomalies?
My guess based on the very limited press release and some aerospace experience:
Because it's only a 4 day slip, they'll start with simulations to look for any systems that may have been damaged. If they don't see any issues assembly and launch will continue.
If simulations show that this may have damaged a component, we'll see a longer delay as they begin replacing systems.
RUAG makes most of the "payload adapters" that integrate the satellite onto the top of a rocket. Based on the size of JWST they're probably using a 1666 model payload adapter.
RUAG's clamp-bands and separation system have a very well characterized Shock Response Spectrum [0, pg. 4], which would have been accounted for in the design of JWST. But the spacecraft and satellite usually separate after the initial shock of the explosive bolt and clamp-band de-tensioning, so they may need to investigate any coupling or recontact of the clamp-band.
If we're lucky they had some accelerometers running during integration and can review actual response data.
Both of those are more shock events - there is a one-time impulse of energy into the system. The main difference here would be damping - the punching bag absorbs and dissipates the energy readily via internal friction of whatever the stuffing material is. A metal bell has orders of magnitude lower internal damping so it sits there and vibrates at its natural frequencies, emitting sound as it pushes against the air which slowly bleeds off energy due air resistance.
Random Vibration is usually characterized by a constant input of energy to the system at a range of frequencies - rocket motors, engine rpm + cylinder firing frequency, road noise, etc.
A somewhat imperfect analogy that I've heard is random vibration is analogous to playing white noise/pink noise/etc through a speaker. Shock is analogous to the horrific crackling and popping you get when you plug a guitar into an amplifier that's already on.
Right, in design shock and vibration are basically 2 separately requirements, analyzed and tested separately. Sometimes by analysis it is shown that vibration test levels "cover" the shock requirement, but a spacecraft that is the first of its type, and with many mechanisms is likely to undergo discrete shock testing.
Minor addition to your reply, design typically avoids similar natural frequencies to the launcher, but the design needs to be compatible with the loads resulting from vibration as
It sounds like the clamp band was accidentally released just after installation, thankfully in a position that didn't drop the spacecraft somehow. The input shock is more like 1000G for these things, usually the largest source of shock for a spacecraft. Of course, it's designed to be released by the same mechanism, although it would not be designed to be released an unlimited number of times. I would guess they already know the components which are critical for shock, and will inspect and/or function test those before proceeding. Since it was an unplanned event, even documenting that it happened and getting everyone to agree to proceed based on the spacecraft having been designed for that input shock, takes at least a few days.
It does not sound like a larger than design shock was somehow applied and they have to check everything at spacecraft level analysis. That would take at least several weeks.
Maybe Elon is right, and instead of spending tens of billions of dollars and decades on one bulletproof piece of jewelry, maybe we should just launch 20 cheaply built ones with less redundancy and hope a couple make it through.
Think monolith mainframe versus cluster of commodity hardware.
There was a great article about this recently on HN, about how Spaceship will change everything with it's extremely cheap price per kilo and how NASA is not ready for it.
This is not Elon’s idea, there was much clamoring in the scientific community to spread the research dollars instead of having one flagship suck up all the funding. It’s a big gamble to be able to see what JWST is designed to see, no number of smaller projects would match it, whether it was the best use of our taxdollars remains to be seen
> Maybe Elon is right, and instead of spending tens of billions of dollars and decades on one bulletproof piece of jewelry
Musk is pretty much socially incompetent, proven so many times. On this topic he discounts the fact that the type of individuals needed to do this job have plenty of offers and nobody wants to work on something which will blow up and has very low odds of making it. Especially smart people who bathe in the concept of opportuinity cost with regards to their time and carrer.
> Do you think that 20 cheap telescopes give the same benefit as one great one?
Arguably, it could give greater benefit.
Time on the Hubble (now) and JWT (eventually) is very carefully controlled. Observation campaigns have to be scheduled far in advance since the camera can only point in one place at one time. Having twenty lower-spec space telescopes would allow twenty research groups their own simultaneous time on the platform.
You seem to think that the same research would be possible on the 20 cheaper ones.
Having a lower quality instrument would likely mean that a lot of the observations would not be possible and that the things that would still be possible would have worse data because the instrument would be worse.
Its easy to say "Just build two of them" or to say the telescope shouldn't have so many single points of failure.
To that I say that a single copy is the only reasonable option. JWT is essentially a prototype and a product in one.
The first to employ such a complex radiation shield,
The first to have berrylium mirrors (AFAIK),
The first to orbit L2.
A copy would inherit any unknown problems with the first telescope, which is especially bad if you've read how many times the JWT has run overbudget and ran into issues. You're better off launching the prototype product to learn what will go wrong and go from there.
The lack of doubles in space shows how exiting the field is because I'd rather have novel prototypes than another hubble.
Yes Spirit and Oppy were made together and it's the same rover design. Also relevant for this discussion due to circumstances one of them outlasted the other 8 years. Imagine if we only had sent Spirit.
Good point. But doing two concurrenct desing would allow as backups for failures not in the telescope itself. For things like lauch, assembly, handling, etc.
If everything goes as planned, having two observatories sure would be appreciated by the scientific community.
Also, I assume, most the costs are associated with the project and setup of fabrication.
Without a complete and honest accounting of the risk profile, which is something NASA and its contractors have historically not been great at, it's hard to say whether test/backup hardware would have been warranted.
IMO "build two, launch one and see what happens" was absolutely not the right approach here. But the goal of the program is to get a functional JWST instrument into position, and if there's any relatively (wrt. the entire program) cheap way they could have substantially de-risked that outcome—by tooling up for a possible second production run, flying a "dumb" article to test the deployment procedure in situ, or whatever else—it should have been seriously considered.
Frankly, this sort of judgment is something I don't trust NASA to get right in practice, through no major fault of their own. Congress runs NASA like they run the rest of the country: setting them up to fail, blaming them when they inevitably do, withholding realistic funding for congressionally mandated programs while simultaneously helping their military-industrial cronies wring out every last dollar. The only way to get anything at this scale done is to lowball up front and beg for more money later.
"NASA's follow-on instrument to the wildly successful Hubble Space Telescope was originally due to launch about a decade ago"
I don't think anything else could better express just how bad the government is at deploying capital than the sentence above. Say what you will above SpaceX, but what they've been able to achieve in the time that this launch has been DELAYED is astounding.
The question was never how to launch it or when but building and testing the telescope. SpaceX has no proven or even theoretical skill in building telescopes.
Building a giant space telescope where everything has to work right the first time is hard. Building a capital-intensive project like Webb on a NASA budget cycle is hard. I find the idea that everyone at NASA is a bunch of fools a pretty tiresome theme in these discussions.
Yeah, as others have said: Is there anything in the private sector remotely close to the JWST in its capability and planned lifespan? If so, why hasn't it launched, or why hasn't another nation developed it?
> I don't think anything else could better express just how bad the government is at deploying capital than the sentence above.
Not really. This is a great expression of how hard it is to deploy capital in a one-off manner, to a strict budget.
Contrast this space telescope with the series of Earth-observation satellites launched by the National Reconnaissance Office. By amortizing development costs over at least half a dozen launches, the NRO keeps putting up classified payloads.
SpaceX's success isn't exactly organizational genius, it's that its plan from the outset has been to rapidly iterate with the expectation that the first half-dozen of anything will fail in interesting ways. Were JWT being built to a SpaceX standard, there wouldn't be one telescope -- there would be a dozen.
Your point that government is bad at deploying capital, as evidenced by the fact that SpaceX accomplished something entirely unrelated during the delays on an incredibly ambitious project that no organization has ever undertaken?
Yes, telescopes are not rockets, and they’re not satellites. But to say they’re entirely unrelated is silly. Even if a reusable rocket is 10x easier to build, it’s obscene to justify a 10 year delay.
I don't believe the JWST effort is comparable to anything SpaceX is doing.
Falcon 9 first optimized for manufacturing cost for a well-proven category of product. Pretty much every aspect of orbital multi-stage rocket design is well explored, and they can surely hire people with experience in it. Optimizing for manufacturing cost is also an activity that many people have experience in. It's an innovative and risky business strategy, but nothing being done is physically or technologically risky - the only risk is they may not be able to optimize costs enough, which is only a risk to the business.
For the next innovation, first-stage landing, it was also very low-risk to their core business. If they screw up the landing stuff, then the first stage blows up somewhere, but they were all going to do that anyways, no risk to the primary payload. This let them iterate on it rapidly for very low cost and risk.
Building a brand-new space telescope of an original design with a dozen new and revolutionary technologies is a whole different ballgame. There's a ton of unknown unknowns that nobody can anticipate ahead of time, only one try per device, and it's either total success or failure. It shouldn't be too surprising that it overran expected costs and deadlines so far. It's more a statement on the Government contracting process that attempts to treat building revolutionary technology just like building the 10 millionth Jeep.
Would it be possible to build an autonomously-assembling modular space telescope, launched piecemeal, with each piece rendezvousing at L2 and bolting on to the previous pieces? The idea being to lower the launch stakes.
> Let's be honest, words like "incident," "sudden," and "vibration" are not the kinds of expressions one wants to hear about the handling of a delicate and virtually irreplaceable instrument like the Webb telescope.
This reminded me of the terms the kerbal space program community use like "Rapid Unplanned Disassembly (RUD)" and "Lithobraking" (crashing into the ground thus "braking" and often breaking, an usually unwanted alternative to aerobraking).
I wouldn't be surprised to see the KSP community use Sudden Vibration Incident as a euphemism for what happens when you exit time-warp and sometimes the whole spaceship turns into a vibrating jelly.
146 comments
[ 3.1 ms ] story [ 210 ms ] threadMake them agree on the one who should shoulder responsibility for a possible problem, to crash the process.
But hey, we need to launch, right?
E: Hey we're pretty sure this is going to kill people.
M: Are you sure? Maybe we can just do it while thinking really positive thoughts.
Meanwhile one of the managers responsible for giving the 'go signal' still blames the engineers. "Hey, if it was unsafe the engineers should have said so." Ignoring that he had been bullying them for hours to shut up about how it was unsafe.
The dramatization of the event in one of the documentaries stands out to me. "I'm going to go around the room and if anyone knows that this thing is definitely unsafe, then they need to speak up." WRONG. You go around the room and make everyone explicitly say it's safe. If anyone refuses you don't launch. They might seem like logical duals, but there is a huge difference.
This problem was a management problem.
Of course they will come out and say "just a short delay" immediately, but who knows what the actual delay gets to when they really start looking at it. Besides, how can they really know that it is okay without taking it down and putting it through all of the motions? A bent arm in a manner not visible while folded in launch config but prevents it from unfolding after launch is a very dumb reason to not have this thing work.
https://youtu.be/aXQ2lO3ieBA
Have you considered what you are proposing?
They would have to send a spyglass if they made it rugged piece of technology not most advanced space telescope in history of humanity.
Any kind of extra shielding, extra rigidity takes away from the weight budget of the project.
It’s a pet peeve of mine but whenever I read “Maybe, just maybe” I instantly dislike whatever comes next. It’s an indicator that the post is sardonic and dramatically over simplified.
The real reason why the JWST is so delicate is because it’s a large satellite that had to be crammed into a tiny fairing. That means it has to be folded up like the worlds most intricate origami. All those moving parts, coupled with incredibly thin and weight saving components means that deploying it successfully is tremendously fraught with complexity. If we had a wider diameter rocket, JWST wouldn’t be nearly as complicated. But that simply wasn’t the case in 1997 when it was first designed.
Not sure which configuration this launch entails but the telescope itself weighs 6,500 kg, so that sounds like a lot of energy to account for.
Ariane 5 user's manual, PDF: https://www.arianespace.com/wp-content/uploads/2015/09/Arian...
It's always agrressive against the users. The users are the poor saps that have to accept whatever the devs have droped on them. The users have no say other than suggest it doesn't work, doesn't promote efficient workflows, whatever. It's then up to the devs to fix, but they feel like they've delivered according to specs so tough. The User will Accept Whatever from the all powerful dev team. How about User Acceptability Testing. At least it then makes it sound like the devs are going to give a damn about the users feedback.
I think it just goes to show that nobody likes creating documentation.
But it's odd to see that on word documents themselves.
Either way, it's chuckles all the way down.
https://www.ulalaunch.com/docs/default-source/rockets/atlasv...
Though do note:
> This user’s guide is intended for pre-contract mission planning and for understanding SpaceX’s standard services. The user’s guide is not intended for detailed design use. Data for detailed design purposes will be exchanged directly between a SpaceX customer and a SpaceX mission manager.
dozens if not hundreds of HNers begin poring thru the manual
If the HARD requirement is "you have this much mass and volume to make a thing to a particular confidence score" you make the damn thing and hope you are on the correct side of the dice.
No single engineer gives a shit about the entire system because no single person is equipped with the human ability of comprehending each individual part that has whatever confidence rating that is required.
The solar shield is like rolling out five layers of tightly packed tinfoil from several points without tearing. Claiming that you would do away with single points of failure if you were the one designing it is too easy to say on HN
There's also the whole "what if we could avoid mass and volume constraints" tangent, that was discussed in the yesterday's Starship article, but it's more of a theoretical consideration for now.
So, what I learned is that single points of failure on initial deployment -- which by definition only has to run once -- is a drastically different cost vs risk calculation than something that happens during normal operations.
You are right that a magical rocket could solve many of NASAs problems (just launch ten Webbs and use the ones that don't work as sunshades for the rest!) but I think NASA is being pretty practical in not waiting for one to be invented.
What was the original budget for JWST? Where's it at now? From your assessment and the numbers, it sounds like there's not just double but triple redundancy now.
Still though, if everything works and we learn more about the universe, I think it'll have been worth it.
(Ok, "weeks" isn't quite applicable in that case... Unless you're counting the time for updates to propagate through all upstream dependencies.)
Does the JWST just operate at a different level of orbit or something?
(edit: don't just single out Russia)
https://en.wikipedia.org/wiki/Lagrange_point
Also, the JWT is at L2, 1.5m km from earth, whereas ASAT tests generally occur in LEO. Cosmos 1408 as you mention orbited at around 500km.
Although I’m certain we could blow up the JWT if we really really wanted to.
Maybe you should name and shame everyone that done this same thing, even if you don't remember similar actions the comments on the Russia storied remind us that other assholes also did the same, some even blew up nukes in orbit causing malfunction to satellites.
Yeah, it is depression to think that lot of things are at the mercy of a few giants whims.
Here a couple of Wall Street journal commentary (AKA opinion) pieces on the topic. They all talk about different aspects of it, the first being probably the most interesting around your point of view.
https://www.wsj.com/articles/a-space-treaty-to-stop-debris-a...
https://www.wsj.com/articles/vladimir-putins-space-front-rus...
https://www.wsj.com/articles/space-armageddon-and-putins-thr...
As always, you can plug these into http://archive.is to bypass the paywall.
It is not rated for being dropped.
While having a Starship sized fairing to put JWST in for launch would probably reduce some packaging complexity, I'm not sure the launch vehicle is the major source of cost in this case.
-- S. R. Hadden, Contact (https://www.youtube.com/watch?v=Et4sMJP9FmM)
The US isn't a serious nation any longer. The US is increasingly becoming a clown show. The response you'd get if you suggested to consider building two of them - for the reasons you mention - is that people would freak out (two boondoggles, $20 billion in cost, decades of time; that would be the flailing, screeching response).
And rightfully so.
What's the ROI? The hard limit for stuff which is not aimed at producing an improvement in quality of life of taxpayers should be in the dozens of millions , not billions.
Quality of life improvement can be delayed 40 or 50 years, but there should be at least a path.
James Webb Space Telescope has no path, nor does Hubble, nor does the ISS, nor do the Mars missions (private or public), and neither did the Apollo missions as the program was shut down for bankruptcy essentially.
Had the Soviet done the same we'd be laughing at their juvinile understanding of economics, ROI and ROE.
On the other hand NASA mission to collide with an asteroid has a path to save us all , ITER has a path (no pun intended) , mRNA research has a freaking highway at this point, biotech also has a path, even China's bridges to nowhere to pump GDP have a path sometime in the future.
Seems like a good pitch to fund National Labs instead.
National Labs != space.
Unless you have a clear and defined QOL plan then space has proven to be a money sink for the longest time.
The Apollo mission didn't have any clear plan, the Voyager-1-2 missions didn't have any clear plan, Mars missions won't have any clear plan, the ISS didn't have any clear plan, the space shuttle didn't have any clear plan.
Matter of fact it's the opposite of what you claim, if you leave space in the hands of military maybe some QOL trickle down will happen (GPS, Sat communications, ICBMs keeping us safe).
If you leave it in the hands of wide eyed individuals who want to discover exoplanets and clusters which we'll never reach then you'll just throw money down the sink.
Unless we can change our essential nature, I don't see any sustainable path where human civilization stays confined to Earth forever. Those doomsday weapons "keeping us safe" will inevitably be used, especially when it becomes clear that we're utterly trapped.
If you gave projects like Webb the kind of budget that now gets eaten by human space flight, you could have telescopes filling half the sky. But NASA has to work within the really severe funding constraints imposed by the white elephant that is ISS.
It sounds to me like this is more of an unintended shock event (see zeeb's response below) that caused an oscillation at the natural frequency ("vibration") but they tortuously avoided using the word "shock".
This would result in a given shock response spectrum [1], which effectively can amplify the input if it hits any part's natural frequency - e.g. a 1G input shock could become a 5G response somewhere in the satellite. I assume this wasn't accounted for in design so they'll determine if this could have damaged any onboard systems.
Launch vibration has a well known Power Spectral Density [2] which effectively says how much energy is being input at every frequency. If they design the system so no parts have a natural frequency at a high-energy peak it's safe.
Shock and vibration are closely related, but very different in practice. It's all about natural frequencies and what frequencies the energy is being input at.
1) https://en.m.wikipedia.org/wiki/Shock_response_spectrum
2) https://vru.vibrationresearch.com/lesson/what-is-the-psd/
Because it's only a 4 day slip, they'll start with simulations to look for any systems that may have been damaged. If they don't see any issues assembly and launch will continue.
If simulations show that this may have damaged a component, we'll see a longer delay as they begin replacing systems.
RUAG's clamp-bands and separation system have a very well characterized Shock Response Spectrum [0, pg. 4], which would have been accounted for in the design of JWST. But the spacecraft and satellite usually separate after the initial shock of the explosive bolt and clamp-band de-tensioning, so they may need to investigate any coupling or recontact of the clamp-band.
If we're lucky they had some accelerometers running during integration and can review actual response data.
0) https://www.ruag.com/system/files/media_document/2019-03/PLE...
Random Vibration is usually characterized by a constant input of energy to the system at a range of frequencies - rocket motors, engine rpm + cylinder firing frequency, road noise, etc.
A somewhat imperfect analogy that I've heard is random vibration is analogous to playing white noise/pink noise/etc through a speaker. Shock is analogous to the horrific crackling and popping you get when you plug a guitar into an amplifier that's already on.
Minor addition to your reply, design typically avoids similar natural frequencies to the launcher, but the design needs to be compatible with the loads resulting from vibration as
It sounds like the clamp band was accidentally released just after installation, thankfully in a position that didn't drop the spacecraft somehow. The input shock is more like 1000G for these things, usually the largest source of shock for a spacecraft. Of course, it's designed to be released by the same mechanism, although it would not be designed to be released an unlimited number of times. I would guess they already know the components which are critical for shock, and will inspect and/or function test those before proceeding. Since it was an unplanned event, even documenting that it happened and getting everyone to agree to proceed based on the spacecraft having been designed for that input shock, takes at least a few days.
It does not sound like a larger than design shock was somehow applied and they have to check everything at spacecraft level analysis. That would take at least several weeks.
Think monolith mainframe versus cluster of commodity hardware.
There was a great article about this recently on HN, about how Spaceship will change everything with it's extremely cheap price per kilo and how NASA is not ready for it.
Musk is pretty much socially incompetent, proven so many times. On this topic he discounts the fact that the type of individuals needed to do this job have plenty of offers and nobody wants to work on something which will blow up and has very low odds of making it. Especially smart people who bathe in the concept of opportuinity cost with regards to their time and carrer.
Like if I gave you 20 shitty 360p cameras to shoot a movie would it look as good as if I gave you one Arri Alexa?
Arguably, it could give greater benefit.
Time on the Hubble (now) and JWT (eventually) is very carefully controlled. Observation campaigns have to be scheduled far in advance since the camera can only point in one place at one time. Having twenty lower-spec space telescopes would allow twenty research groups their own simultaneous time on the platform.
Having a lower quality instrument would likely mean that a lot of the observations would not be possible and that the things that would still be possible would have worse data because the instrument would be worse.
Please, make that argument. Explain to us how 20 - hell let's be generous and say 1,000 - worse telescopes can achieve what one great one can.
To that I say that a single copy is the only reasonable option. JWT is essentially a prototype and a product in one. The first to employ such a complex radiation shield, The first to have berrylium mirrors (AFAIK), The first to orbit L2.
A copy would inherit any unknown problems with the first telescope, which is especially bad if you've read how many times the JWT has run overbudget and ran into issues. You're better off launching the prototype product to learn what will go wrong and go from there.
The lack of doubles in space shows how exiting the field is because I'd rather have novel prototypes than another hubble.
There are some doubles though. Voyager 1 & 2, Spirit and Opportunity…
I think the spy hubbles are better examples.
Can you imagine the added complexity for the Hubble design for it to work as space observatory or earth observatory.
I wonder if the he Hubble was myopic because of this.
If everything goes as planned, having two observatories sure would be appreciated by the scientific community.
Also, I assume, most the costs are associated with the project and setup of fabrication.
IMO "build two, launch one and see what happens" was absolutely not the right approach here. But the goal of the program is to get a functional JWST instrument into position, and if there's any relatively (wrt. the entire program) cheap way they could have substantially de-risked that outcome—by tooling up for a possible second production run, flying a "dumb" article to test the deployment procedure in situ, or whatever else—it should have been seriously considered.
Frankly, this sort of judgment is something I don't trust NASA to get right in practice, through no major fault of their own. Congress runs NASA like they run the rest of the country: setting them up to fail, blaming them when they inevitably do, withholding realistic funding for congressionally mandated programs while simultaneously helping their military-industrial cronies wring out every last dollar. The only way to get anything at this scale done is to lowball up front and beg for more money later.
Still under one month? At least the slope is still less than one[1]
[1]https://xkcd.com/2014/
I don't think anything else could better express just how bad the government is at deploying capital than the sentence above. Say what you will above SpaceX, but what they've been able to achieve in the time that this launch has been DELAYED is astounding.
Not really. This is a great expression of how hard it is to deploy capital in a one-off manner, to a strict budget.
Contrast this space telescope with the series of Earth-observation satellites launched by the National Reconnaissance Office. By amortizing development costs over at least half a dozen launches, the NRO keeps putting up classified payloads.
SpaceX's success isn't exactly organizational genius, it's that its plan from the outset has been to rapidly iterate with the expectation that the first half-dozen of anything will fail in interesting ways. Were JWT being built to a SpaceX standard, there wouldn't be one telescope -- there would be a dozen.
Falcon 9 first optimized for manufacturing cost for a well-proven category of product. Pretty much every aspect of orbital multi-stage rocket design is well explored, and they can surely hire people with experience in it. Optimizing for manufacturing cost is also an activity that many people have experience in. It's an innovative and risky business strategy, but nothing being done is physically or technologically risky - the only risk is they may not be able to optimize costs enough, which is only a risk to the business.
For the next innovation, first-stage landing, it was also very low-risk to their core business. If they screw up the landing stuff, then the first stage blows up somewhere, but they were all going to do that anyways, no risk to the primary payload. This let them iterate on it rapidly for very low cost and risk.
Building a brand-new space telescope of an original design with a dozen new and revolutionary technologies is a whole different ballgame. There's a ton of unknown unknowns that nobody can anticipate ahead of time, only one try per device, and it's either total success or failure. It shouldn't be too surprising that it overran expected costs and deadlines so far. It's more a statement on the Government contracting process that attempts to treat building revolutionary technology just like building the 10 millionth Jeep.
This reminded me of the terms the kerbal space program community use like "Rapid Unplanned Disassembly (RUD)" and "Lithobraking" (crashing into the ground thus "braking" and often breaking, an usually unwanted alternative to aerobraking).
I wouldn't be surprised to see the KSP community use Sudden Vibration Incident as a euphemism for what happens when you exit time-warp and sometimes the whole spaceship turns into a vibrating jelly.